[0005] This disclosure presents a phase change memory cell structure having a phase change material with a switching current path that is confined to a region smaller than the potentially switchable volume of the entire phase change material. The disclosure further presents a method of forming the phase change memory cell structure. The structure comprises a phase change memory cell having a main body of phase change material. The minimum width of the main body of phase change material is limited only by the minimum dimensions attainable using optical lithography techniques (i.e., minimum lithography dimensions). Additionally, the memory cell structure has a narrow channel of phase change material connecting to the top electrical contact. This narrow channel filled with phase change material provides a confined region for the switching current path. Thus, the phase change material in the main body of the cell does not switch in response to an applied current. In other words the main body of phase change material effectively becomes a non-switching portion of the phase change material. The phase change material in the narrow channel does switch in response to an applied current. In other words the narrow channel of phase change material becomes a switching portion of the phase change material. The width of the narrow channel is less than the width of the main body and, thereby, less than minimum lithographic dimensions. Lastly, an embodiment of the phase change memory cell structure further isolates the main body / non-switching portion of the phase change material by providing a space between the phase change material and the memory cell walls. The space may comprise, for example, a vacuum or a gas (e.g., a low pressure argon gas). The space allows for the expansion and contraction of the phase change material and limits heat dissipation from the phase change material.
[0007] In another embodiment, the phase change memory cell also comprises a phase change material comprising a main body of phase change material (i.e., a non-switching portion of the phase change material) in series with a narrow channel filled with phase change material (i.e., switching portion of the phase change material). More particularly, the phase change memory cell comprises a first metal contact with a first liner (i.e., a bottom electrical contact) that contacts the main body of phase change material. An insulator layer is positioned adjacent the first liner and an etch-resistant layer is positioned adjacent the insulator layer. The etch-resistant layer (e.g., an HF etch-resistant metal) comprises a top surface and a bottom surface. The etch-resistant layer also has a narrow channel that is filled with the phase change material and extends from the top surface to the bottom surface. The first insulator layer has a cavity that extends between the bottom surface of the etch-resistant layer to the first liner layer. The narrow phase change material-filled channel opens into the cavity and the main body of phase change material is formed in the cavity. A second liner and second metal contact (e.g., top electrical contact) is positioned adjacent the switching portion of the phase change material in the narrow channel. A minimum width measurement of the narrow channel of phase change material (i.e., switching portion) is less than any width measurement of the main body of phase change material (i.e., non-switching portion). Specifically, the minimum width measurement of the channel comprises a less than current state-of-the-art minimum lithographic dimension. In this particular embodiment, the main body of the phase change material (i.e., non-switching portion) does not fill the entire cavity, but rather extends pyramid-like from the liner towards the narrow channel. Additionally, the narrow channel may be tapered towards the main body of phase change material such that the channel is wider near the top electrical contact than it is near the main body. Thus, a space (i.e., gap) remains around the non-switching portion. This space may be a vacuum space or a gas-filled space (e.g., a low-pressure argon-filled space). The space isolates the non-switching portion from the first insulator. The space allows the non-switching portion to expand and contract and limits heat dissipation.
[0008] One embodiment of a method of forming a phase change memory cell comprises forming a multi-layer stack disposed within an insulator. The stack should comprise a first metal contact layer, a first liner layer adjacent the first metal contact, a phase change material layer adjacent the first liner layer, and a nitride layer adjacent the phase-change material. Then, a first recess is etched through the insulator to the nitride layer. The first recess is etched by first depositing a photo-resist layer over the insulator. The photo-resist layer is lithographically patterned such that the first recess preferably has a width with the minimum lithographic dimension possible. Using the pattern, the first recess is etched and the photo-resist layer is removed. A spacer material layer (e.g., silicon oxide, silicon nitride, aluminum oxide, etc.) is deposited to form an approximately conformal layer over the insulator and on the nitride layer exposed in the first recess during the etching process. A second recess, having a minimum width with a less than minimum lithographic dimension, is then etched through the spacer material and the nitride layer in the first recess to expose the phase change material layer. The second recess is anisotropically etched through the spacer material layer and the nitride layer in the first recess, thereby, forming the second recess. The second recess may be tapered towards the main body of phase change material such that the resulting channel that is formed narrows towards the phase change material layer. Once the second recess is etched, an additional layer of the phase change material is deposited over the insulator and onto the spacer material layer and the first phase change material layer in the second recess. The additional phase change material layer is polished. Then, a second liner layer can be formed over the additional phase change material layer and finally a second metal contact layer can be formed on the second liner layer. The resulting structure comprises a narrow channel of phase change material in series with a wider main body of phase change material between two contacts. The density of current passing between the electrical contacts through the narrow channel is increased relative to the main body and thereby restricts the switching volume of the phase change material to within the narrow channel.
[0009] Another embodiment of a method of forming a phase change memory cell comprises forming multi-layer stack disposed within an insulator. The stack should comprise a first metal contact layer, a first liner layer adjacent the first metal contact layer, and a phase change material layer adjacent the first liner layer. A first recess is etched through the insulator to the phase change material layer. Etching the first recess is accomplished by depositing a photo-resist layer over the insulator. The photo-resist layer is lithographically patterned such that the first recess preferably has a width with the minimum lithographic dimension possible. After the first recess is etched, the photo-resist layer is removed. A nitride layer (e.g., an aluminum nitride layer, a silicon nitride layer, etc.) is deposited to form an approximately conformal nitride layer over the insulator and on the phase change material layer in the first recess. Then, a spacer material layer (e.g., silicon oxide, silicon nitride, aluminum oxide, etc.) is deposited to form an approximately conformal layer over the nitride layer. A second recess is anisotropically etched such that it has a minimum width with a less than minimum current state-of-the-are lithographic dimension. The second recess is etched through the spacer material layer and nitride layer in the first recess to expose the phase change material layer. The second recess may be tapered towards phase change material layer. An additional layer of the phase change material is deposited over the insulator and onto the spacer material and the phase change material layer in the second recess. The additional phase change material layer can be polished. Then, a second liner layer can be formed over the second phase change material layer and finally a second metal contact layer can be formed on the second liner layer. The resulting structure comprises a narrow channel of phase change material in series with a main body of phase change material. The density of current passing between the electrical contacts through the narrow channel is increased relative to the main body and thereby restricts the switching volume of the phase change material to within the narrow channel.
[0010] Another embodiment of a method of forming a phase change memory cell comprises forming a multi-layer stack on a silicon wafer. The multi-layer stack should comprise a first metal contact layer isolated from the silicon wafer by an insulator, a first liner layer adjacent the first metal contact layer, an oxide layer adjacent the first liner layer, and a hydrofluoric acid (or buffered hydrofluoric acid) etch-resistant layer adjacent the oxide layer. A first recess is etched through the hydrofluoric acid (HF) etch-resistant layer to the oxide layer. Etching the first recess is accomplished by first depositing a photo-resist layer on the HF etch-resistant layer. The photo-resist layer is lithographically patterned such that the first recess preferably has a width with the minimum lithographic dimension possible. The recess is etched and the photo-resist layer is removed. A nitride layer (e.g., a silicon nitride and an aluminum nitride) is deposited to form an approximately conformal layer over the HF etch-resistant layer and the oxide layer in the first recess. A second recess is directionally etched through the nitride layer in the first recess and through the oxide layer to expose the first liner layer such that the second recess has a width with a less than minimum lithographic dimension in the first direction. Then, the oxide walls of the second recess are isotropically etched from the second recess to form a cavity between the HF etch resistant layer and the first liner layer. Thus, the second recess becomes channel opening into the cavity and may be tapered as it nears the cavity. A phase change material layer is deposited into the channel and the cavity. A sputtering process is preferably used to deposit the phase change material such that such that a main body of the phase change material extends pyramid-like from the first liner layer to the narrow channel and a portion of the phase change material also fills the narrow channel. Any width measurement of the main body of phase change material in the cavity is greater than a minimum width measurement of the narrow channel filled with phase change material. Furthermore, a space remains in the cavity around the main body phase change material. This space isolates the phase change material from the oxide layer. Isolating the first phase change material allows it to expand and contract and also limits heat dissipation. After the phase change material layer is deposited, it is polished. Then, a second liner layer can be formed over the phase change material layer and finally a second metal contact layer can be formed on the second liner layer. The resulting structure comprises narrow channel of phase change material in series with a wider main body of phase change material. The density of current passing between the electrical contacts through the narrow channel is increased relative to the main body and thereby restricts the switching volume of the phase change material to within the narrow channel.